red-white-project-progress-ppt
Pass
Audited by Gen Agent Trust Hub on Jul 1, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: Indirect prompt injection surface identified. The skill ingests untrusted user content (such as project progress and business reviews) and interpolates it directly into a prompt template for the
image_gentool without sufficient isolation. - Ingestion points: User-provided text segments used to populate the image generation prompts as described in
SKILL.md. - Boundary markers: The prompt template lacks robust delimiters or explicit instructions to ignore commands embedded within the user's data, which could lead the model to follow instructions hidden in the input.
- Capability inventory: The skill is configured to automatically call the
image_gentool to produce images based on the generated prompt. - Sanitization: No input validation or sanitization is mentioned to filter out potential adversarial commands within the user's input before prompt interpolation.
Audit Metadata